The need for large scale computing systems continues to increase as enterprises are generating and consuming greater amounts data. This need is especially true for enterprises providing access to data by large numbers of concurrent user sessions. As computing systems scale to handle the load required for processing numerous concurrent user sessions, so does the complexity of such systems. This complexity can be due, in part, to the interrelationship amongst various services executed by the system in response to user requests and the like. Execution of these services can require significant processing resources and given their related nature, crashed processes of one user often affect processes of other users (i.e., the other processes can also crash).